Multiplicative Contractions, Additive Recoveries: Functional-Form Restrictions on Risk Exposure Dynamics

Abstract: We test a regime-conditional functional-form restriction on aggregate risk-exposure dynamics implied by VaR-constrained intermediary models: exposures contract multiplicatively when capital constraints bind and grow additively (level-independent) when slack. The contraction half follows from binding VaR constraints (Brunnermeier and Pedersen 2009; Adrian and Shin 2010; He and Krishnamurthy 2013). The additive-rebuild prediction is derived under constant-rate capital replenishment; we test the joint restriction on FINRA monthly margin debt (1997-2026). Two findings. First, regime-interacted regression of detrended margin growth on lagged level (T=350 months) yields calm slope -0.040 (p=0.082, additive) and stress slope -0.205 (p<0.001, multiplicative); Wald test on regime x level interaction rejects equal dependence (p=0.0016). Second, the restriction implies drawdown-recovery duration ratio increases with crash depth. On 73 S&P 500 episodes (1950-2026), Cox model gives depth coefficient -13.75 (p<10^{-7}): 75% lower recovery hazard per 10pp deeper drawdown. Continuous-depth regression yields beta=1.22 (p=0.047); beta=1.59 (p<0.001) excluding 1980-82 Volcker. Median duration ratio for crashes >30% is 3.1x; replicates across eight other equity indices. Calibrated Heston, Markov-switching, and block bootstrap nulls match price-level duration asymmetry but lack an exposure state variable, so cannot speak to the regime-conditional flip on direct exposures. We do not claim the exposure test identifies the intermediary mechanism: FINRA margin debt is a noisy proxy. We claim only that the regime-conditional functional form is a sharper target than return-level moments alone, and confirming it on margin debt is consistent with -- not proof of -- the constrained-intermediary mechanism. A companion test on CFTC weekly speculative positioning is left for future work (Sections 5.2 and F).

• The paper introduces a regime-conditional functional-form restriction where exposures contract multiplicatively in stress regimes and recover additively in calm periods.
• It empirically validates the MC–AR restriction using FINRA margin debt and S&P 500 drawdown-recovery durations, with statistically significant regime-dependent effects.
• The findings have practical implications for risk management policy and intermediary behavior, highlighting limitations of return-only models.

Regime-Conditional Restrictions on Risk Exposure Dynamics

Theoretical Foundation of Multiplicative Contractions and Additive Recoveries

This paper introduces and tests a functional-form restriction on aggregate risk-exposure dynamics, motivated by VaR-constrained intermediary asset pricing models. The central theoretical assertion is the existence of a regime-conditional asymmetry: exposures contract multiplicatively during binding-capital-stress periods, while they grow additively, in a level-independent manner, during calm regimes. The multiplicative contraction follows directly from capital constraints (standard in liquidity-spiral and procyclical-leverage models), where intermediaries adjust positions in proportion to their current size under constraint activation. However, the additive recovery dynamic is newly formalized here, derived under the assumption of constant-rate capital replenishment. This yields a distinctive “MC–AR” restriction: the stress regime induces exposure dynamics of the form $X_{t+1} = \bar\eta_t X_t$ with $\bar\eta_t < 1$, and the calm regime produces $E[X_{t+1} - X_t] = \phi > 0$, $Cov(\Delta X_t, X_t) = 0$.

The analytical core comprises two propositions: one for multiplicative contraction (mechanical from binding VaR constraints) and one for additive replenishment (driven by constant-rate inflows and level-independence). The regime-switch directly informs empirical design: the expected growth rate switches from proportional to additive as regime changes, and the functional form is nested in a regression model, isolating these properties via interaction terms.

Empirical Evidence: Functional-Form Testing on Margin Debt

The paper reports robust empirical evidence using monthly FINRA margin debt (1997–2026), detrended against the secular expansion of market capitalization to isolate cyclical changes. The primary regression model interacts the regime indicator (stress vs. calm, classified by volatility thresholding) with the lagged margin-debt level, producing regime-specific slope estimates for exposure growth.

• Calm periods exhibit a slope of $-0.040$ ($p=0.082$), statistically consistent with additive, level-independent growth.
• Stress periods exhibit a slope of $-0.205$ ($p<0.001$), consistent with multiplicative contraction.
• The regime × level interaction coefficient ($\hat b_S = -0.165$, HAC SE $0.052$) is highly significant ($\bar\eta_t < 1$0).
• The stress slope is approximately 5x larger than the calm slope, rejecting equal level-dependence at the $\bar\eta_t < 1$1 significance level.

The robustness of these results is tested against alternative volatility thresholds, detrending methods, and sub-sample analyses. The negative interaction remains stable across all variants.

Figure 1: Raw FINRA margin debt time series from 1997 to 2026, highlighting stress episodes with sharp proportional contractions and calm periods with approximately constant-rate additive growth.

Price-Level Consequences: Drawdown-Recovery Duration Ratios

A direct implication of the MC–AR restriction at the price level is an asymmetric relationship in drawdown-recovery durations: deeper crashes should take disproportionately longer to recover relative to their contraction phase. This is tested via the S&P 500 daily series (1950–2026), identifying drawdown-recovery episodes according to a maximal-interval algorithm.

The empirical findings:

• For $\bar\eta_t < 1$2 drawdowns, the median duration ratio is $\bar\eta_t < 1$3; for 5–10\% corrections, it is $\bar\eta_t < 1$4.
• In continuous-depth regressions, $\bar\eta_t < 1$5 ($\bar\eta_t < 1$6), $\bar\eta_t < 1$7 ($\bar\eta_t < 1$8) in Cox proportional hazard, indicating a $\bar\eta_t < 1$9 reduction in per-period recovery hazard per 10pp drawdown depth.
• The cross-market replication extends this pattern to eight additional international indices, confirming the magnitude-low recovery coupling in 19 severe-crash episodes.

  Figure 2: Drawdown-recovery asymmetry in the S&P 500, with severe crashes showing much longer recovery durations than contraction, in line with the MC–AR prediction.

Discrimination Against Return-Only Null Models

To assess discriminatory power, calibrated Heston stochastic volatility, Markov regime-switching, and block-bootstrap null models are simulated. Each can reproduce observed price-level duration asymmetry (e.g., Heston median $E[X_{t+1} - X_t] = \phi > 0$0, not rejected against empirical S&P 500 $E[X_{t+1} - X_t] = \phi > 0$1), but crucially, none contain an exposure state variable; thus, they are silent regarding the regime-conditional functional-form flip observed in margin debt. The observed exposure-level asymmetry is logically external to these models.

Policy Implications and Theoretical Significance

The regime-conditional restriction refines the empirical targets available for testing intermediary mechanisms. It provides sharper identification than return-level moments alone, suggesting the necessity of intermediary-resolved exposure data for full mechanism identification. Practical implications include nuanced policy design for stress interventions (graduated margin/capital policies vs. binary circuit breakers) and improved calibration of volatility-regime risk management. The MC–AR restriction is consistent with multiple plausible micro-foundations, such as direct inflows, quadratic adjustment cost, and information acquisition; however, additive replenishment is challenged by asset-proportional inflow models, further underscoring its empirical distinctiveness.

Figure 3: Realized volatility of S&P 500 log returns (1950–2026), with volatility clustering highly concentrated during stress-regime months classified above the 90th percentile.

Limitations and Directions for Future Research

The findings hinge on aggregate proxies (FINRA margin debt). Limitations include heterogeneous intermediary responses, noisy proxy effects mixing leveraged demand and revaluation, and lack of intermediary-level tests. The constant-rate replenishment assumption is robust in several foundation scenarios but remains open to theoretical strengthening. Regime classification is by single-threshold volatility; refinements using granular intermediary data could increase empirical power.

Future directions:

• Intermediary-level tests using filings/prime-broker data.
• Companion tests on CFTC speculative positioning.
• Structural models embedding endogenous capital evolution.
• Time-varying restriction strength assessment as leading indicator for mechanism activation.

Conclusion

This paper formalizes and empirically tests a regime-conditional functional-form restriction (“multiplicative contraction, additive replenishment”) on risk exposure dynamics, derived from VaR-constrained intermediary models. The restriction is confirmed on FINRA margin debt and price-level drawdown-recovery data, and is shown to discriminate against prominent return-level null models. While the results are consistent with constrained-intermediary mechanism operation, intermediary-resolved exposure evidence remains essential. The regime-conditional restriction thus advances empirical testing in asset pricing theory, offering new diagnostic targets beyond return-level statistics.

(2604.23315)